What Top Investors Think of the US Software Apocalypse

So far this year, as the US equity market, especially growth and technology stocks, have tumbled, it has looked as if the air has begun to leak from the much-feared and debated artificial intelligence bubble.

Not so. Many semiconductor-equipment and hardware stocks have risen sharply. Software has been the real casualty. Why? Investors are no longer debating whether AI is useful; they are debating whom it will displace. The argument intensified in February after leading AI vendor Anthropic rolled out new tools that exacerbated an existing fear about the software sector: that AI would substitute rather than assist many coders and their products.

Some of the prominent equity managers with whom Morningstar analysts regularly speak as part of their routine due diligence have shared strikingly different views on which companies have competitive economic moats that can withstand AI’s threats, which ones merely have momentum, and which ones are living on borrowed time. Here’s a sample of some of the more compelling views.

Code Red for Software

“In general, that whole sector is just not worth investing in, at least not at this time,” said Fidelity Investments’ Mark Schmehl about software. He runs two of Canada’s largest innovation-focused strategies from San Francisco. Speaking in March 2026, he said software’s old economics depended on a labor scarcity that no longer exists. For years, the sector enjoyed rich margins because writing and maintaining code required expensive, scarce human talent, Schmehl said. AI makes code easier to write and copy and cheaper to maintain, which means software engineering expertise is no longer a unique, competitive moat. Software companies now look more like commodity utilities, whose weaker pricing power and worse long-term economics deserve lower valuations, he said.

The most important thing is, we probably are at artificial general intelligence. We probably will hit it this year or next. We will create a sentient AI. That is going to change the world in fundamental ways. Folks need to start thinking about what that means for the market and for their professional careers.

Not Every Castle Will Fall

Yet Schmehl doesn’t think every software-tagged company is doomed. He argues that investors have sold the group indiscriminately, creating openings for those who can tell the truly exposed from the merely associated. He draws a distinction between pure software firms and those using software to support a more resilient business. Shopify SHOP, he argues, is an e-commerce platform and payments business that belongs in the second camp.

Mark Casey, manager of American Funds Washington Mutual AWSHX, agrees that some software and related firms are insulated from AI disruption or may even benefit from it. He sees Salesforce CRM as one such baby thrown out with the bathwater. Its products are not just bundles of features, Casey contends. It is deeply embedded in customer workflows, acts as a system of record, and is aggressively integrating AI features into its products. The same logic applies to information technology consultants, such as Gartner IT and Accenture ACN, which investors have treated like “AI roadkill,” Casey said. If AI makes technology choices, systems integration, and corporate implementation more complex, demand for the kind of research, advice, and execution help that these firms provide may rise, not fall, he said.

Brilliant Models, Blurry Moats

And what about OpenAI, Anthropic, and Alphabet GOOGL, the providers, respectively, of popular AI models ChatGPT, Claude, and Gemini? Over the past year, Alphabet’s market value has surged. OpenAI’s March 31 USD 122 billion funding round implied an USD 852 billion valuation. Both OpenAI and Anthropic, which also raised money at enormous valuations, could go public in 2026.

Do these richly valued AI companies competing with similar products have enduring competitive advantages? Dennis Lynch of Morgan Stanley’s Counterpoint Global is not sure. Speaking with me during an event in January, he said that AI was clearly important, but he was unsure of paths to profitability. The products look impressive, but they also look, in many respects, interchangeable. If ChatGPT, Claude, and Gemini all do many of the same things, why would users stay loyal to one? Beyond some memory and workflow stickiness, Lynch does not see a clear answer yet. Further, if or when OpenAI or Anthropic goes public, their financials would be exposed to a much wider audience, which might be healthy or uncomfortable. He suggested that private markets often tolerate growth stories that public markets tend to interrogate more ruthlessly.

AI is obviously a big deal... The hard part, often, though, is trying to find a unique company or a gatekeeper that really benefits from that change. Sometimes things improve technologically or efficiency-wise, and it’s just the end users, customers, individuals who benefit, not necessarily one company.

The Labs May Build It. Others May Cash In.

Echoing some of Lynch’s views, in recent months, GQG Partners has argued that OpenAI has little to keep customers from switching, and its technology could become harder to differentiate as competitors catch up. It thinks OpenAI’s massive valuation rests more on “hopes and dreams” than on sound economics, that headline usage figures overstate the quality of demand, and that the unreliability and slowing performance gains of its models could limit adoption in high-stakes corporate work. In GQG’s telling, even if AI changes the world, the labs themselves may not be the ones who profit most; they could end up as low-margin providers of raw intelligence, while companies like Microsoft MSFT and Salesforce capture the richer economics by weaving AI into existing products and workflows.

Peel one layer or dare to bring out your calculator for a closer look, and OpenAI goes from being the world’s “most valuable start-up” to what we believe is one of the most overvalued and overhyped companies in history.

By contrast, Schmehl is optimistic about Anthropic’s prospects, which explains his portfolios’ positions of around 2% each in the private company. Many other Fidelity funds own both of the AI model vendors.

The Real Fight Is Over the Spoils

None of these investors thinks AI is a fad. All think it will disrupt business models. The key argument is not about whether AI will matter but about who will win. Software is the first large sector where investors have tried to price not just AI’s promise but its power to displace. More reckonings will follow.